I have dreamed the same. Perhaps not a switch per se, but effective isolation, like Gigafoil/fibre, and perhaps a very good clock. The best isolation I have experienced is wifi, despite concerns about the detriment of radio waves.
Often the noise we are talking about is not itself within audible frequencies and so one cannot hear it as background noise etc. This is because the noise can be high frequency, sometimes as high as GHz RF noise and the mechanism for this becoming audible is when it finds its way through to the analogue stage of the DAC and causes intermodulation distortion. This can be audible at even very low levels and manifests as a harshness to treble and emphasis to percussion and plucked notes. Initially it might be mistaken for greater soundstage or ‘air’ or ‘life’ in the music but inevitably it becomes fatiguing with extended listening.
It is just my 2p worth of course but hopefully it is aligned with what others have experienced.
To clarify and confirm …
The ‘noise’ I am referring to is the term commonly used in audio-speak to refer to what seems to be interference. It is not itself audible, but its effects are. Often one does not know it exists unlit it is reduced.
In my experience, things that reduce this are:
- Power conditioning
- Cable design
- ethernet isolation
- equipment design
- tweaks like Synthetic Research ECTs
A close cousin of noise is interference and distortion due to acoustics…
There have been times when I have enjoyed some interference and its distorting effects, as it is perceived in general terms as higher resolution, detail, etc when my system was lacking the ability to convey that level of resolution in a clear and enjoyable manner. In other words, noise can brighten up a dull system. It can also be very unenjoyable.
I don’t enjoy noise any more. It is the enemy of purity.
The noise floor of recording I am referring to is the sound left when there is no audible music or effects … the sound of silences. When that silence is audible the system really has a ‘black’ background and full dynamic range can be enjoyed as well as many other aspects of music.
You did an excellent job laying this out. This is exactly what I was speaking of when I mentioned a lower noise floor. But I think the noise @NickBacon mentioned can contribute to whiter backgrounds, so to speak. This is above my pay grade but what I can say is that efforts to address noise can be extremely rewarding. I will never forget the moment when the Sonore opticalModule landed as this was when my network seemed to no longer be doing noticeable harm. Hardness that I had believed to be intrinsic to certain recordings had vanished. It was a Startech FMC that was causing that harm - and it was upstream of a span of fiber. Folks will tell you that fiber offers perfect isolation. They are wrong about it being perfect, unfortunately.
A few years back on one of the forums, someone shared a way to test your system’s resolving capabilities. A few tracks were named and the test had you listen for some things buried deep in each track. I didn’t do as well as a few others with much better systems. I’ll have to see if I can track down that thread as I’d like to repeat it. My room has improved significantly as has my system since I last did that.
Many recordings, even recent ones, have quite noticeable residual background noise which one can hear being faded up at the start of the track just before the music starts.
Shucks, thanks for the compliment.
Yes, noise makes a less black background for sure. But despite a black background I am presently trying to solve siblence that is far more than I’d expect to be allowed by those who recorded the music. It is more on some tracks than others, so perhaps it is in the recording. On other tracks ir is gradually reducing with more tweaking to do so.
I think fibre probably is totally galvanic. However, i gather things like phase noise and distortion are still created by every device in a network, and this is passed on by fibre. This is why a FMC like optical module has such an impact, and why upgrading its power supply is beneficial.
My guess is this also partly explains why daisy chaining switches is also often beneficial … each step reduces noise and distortion in the next.
So, in short, ethernet isolation transformers and fibre are just parts of the solution to building a truly great network for audio.
Sadly this is true. There is only so much that can be dealt with by EQ and other methods. From what I can hear, recording engineers also try to fade out the tail of sustain at the end of tracks so we can’t hear that residual.
Interestingly, I also gather some recordings seem to make little attempt to reduce this, perhaps to be authentic. Or maybe the studio monitoring system didn’t reveal it.
There is a particular U2 track I know that captures very low frequency of a backing singer ocassionally bumping the mike stand. It is very distracting, and surely a recording engineer would have EQd that out. I can only assume the engineer could not hear it
Ultimately someone is making a judgement … is the noise diminishing the intent and enjoyment of the music? If yes, reduce it. Obviously there is a good dose of subjectivity in that judgement that may not align with the listeners ideals.
And we are left to try to work out what is in the recording and what is in our systems.
1 Like
It’s been almost a month since I got the phoenixnet.
please note that this applies to my system. Your results may vary.
The phoenixnet when added in front of k50 and after the twin sotm only Has a slight effect on the sound. The bass gets better and tighter. But otherwise not a huge change to sound.
I think I may have hit the laws of diminishing returns here.
One thing I noticed is that the phoenixnet doesn’t like the sotm DCBL cable connected from it to k50. It’s robs the sound of a lot of sparkle and attack. Infact there’s a further improvement to sound when I also use a genetic Ethernet cable to connect from sotm to phoenixnet.
My stereo is like this
Router —DCBL7–isocat6–dcbl7–sotm switch1—DCBL7–isocat6–dcbl7—sotm switch2–K50
Also. While testing cables or switches please note that the latest roon versions caches the current song and a bit of the next song as well.
And what if you leave out the second Sotm in this chain? Do you need it for sufficient ethernet outlets? Just curious.
That is interesting experience with the PhoenixNet. From what I gather, it is usually quite a significant improvement, and I assume superior to SoTM, so much so it would be worth rebuilding your network based on that.
If you can use a good grade Cat 6 UTP (short as possible) from router to PNet as a baseline. Then put a SoTM cable from PNet to K50 (no ISOCAT). I have an ISO CAT and it makes a little difference, really just polish/refinement, so leave that out until you get a good foundation.
Once you are familiar with that, add a SoTM switch and another Cat 6 UTP.
Once you are familiar with that, from SoTM to PNet swap out the CAT 6 with SoTM cable.
And so on … then you will hear the real value of each switch and cable combination and decide if you like 2 SoTMs or 1 SoTM + PNet … or if all three is justified.
Incidentally, there have been advances in cables since SoTM Cat 7.
And make sure no SMPS is plugged into the same outlet as any of the network gear.
1 Like
A quick Update on my system since adding the phoenixnet
It’s been about 3 Months now since I added the switch. If you remember my initial impressions while good ,were not spectacular.
I recently started feeling that sotm DCBL Ethernet cables I use throughout my system doesn’t have good synergy with the phoenixnet switch. I decided to give the Furutech lab-8 cables a try to test this theory.
Long story short. Using Furutech between my second sotm switch and k50 has given a serious boost in resolution and instrument separation. Maybe even sense of height. I still use sotm cable between the sotm switches and my virgin media router.
I see a few ppl mention online about phoenixnet not liking certain cables (sotm etc) and working fine with some.
For now , I think my upgrade itch is over. Last night me and the Mrs just sat down listening to music. For hours. She usually doesn’t show much interest. But yesterday she was humming along. It was a good feeling 
1 Like
I have tested with my K50 the following ethernet switches:
- Standard D-Link switch
- Silent Angel Bonn n8 (no difference from standard D-Link)
- EtherRegen (better than D-link but still not a big improvement)
- SoTM with internal Clock (very good audible improvement)
- Melco S100 (similar to SoTM above)
- Paul Pang Quad switch (best one… amazing)
The combination of a ENO filter with those switches brings amazing results improving dramatically the sound quality.
- A Standard D-link + ENO reaches the same kind of performance of a SoTM / Melco;
- A SoTM with ENO gets the level of performance of the Quad from Paul Pang;
- Paul Pang Quad with Eno reached an outstanding sound quality only overcome by the same Switch with the MUON filter.
1 Like
Hi LGA, thanks for sharing… did you also try the eR + ENO as well?
why is roon server so bad then. they sell themselves on bitperfect audio. shocked to hear that a free solution (squeeze) is better. What on earth are roon doing wrong to mess with the sound!
AfterDark Dela S100 fully modified.
BLOWN AWAY
Bit perfect playback is a problem that was solved long ago. What we are left with is dealing with sources of noise.
When an application is very active, more noise is generated in the computer. Roon keeps very busy as it is doing a lot of communication back to Roon’s servers. This is how they keep the UI up to date with the latest information about artists and new music. Squeeze server doesn’t do this as it only retrieves additional information when you add an album to your library. It’s important to keep in mind that Roon is not really a standalone application. It’s more of a service, which is why there is subscription pricing.
On the plus side, Roon has improved dramatically as far as how chatty it is. It used to be that my Roon server was always listed as being the busiest device on my network. That changed with one of the 1.8 releases. As I mentioned though, Roon is really a service, so the sound quality can fluctuate from release to release. It’s always going to be busier than a standalone application like Squeeze.
5 Likes
Does a device like the PhoenixNet actually ‘clean’ the feed and remove any noise from it? Because in my experience crap in = crap out. If a dirty noisy signal comes into the house, and then into the broadband providers router, how can sending that dirty signal into an audiophile switch improve things on the other end if it’s already dirty?
@tommytwotimes If you read the info on the Innuos web site about the PhoenixNET I think it makes it clear that it does filter noise. When I tried the PhoenixNET and various other switches, the PhoenixNET was the only one that made my system sound teh same whether or not it was connected to the network. In other words if I set a local file on my K50 playing and then pulled the ethernet cable from the rear of my K50 that sound did not alter if I was using the PhoenixNET. With all the other switches I tried there was a difference when I pulled the ethernet cable.
There are many areas of the digital playback chain that suffer from noise overlaid on the digital signal. Unfortunately just relying on galvanic isolation of a DAC is not sufficient to block the noise because that isolation is rarely (if ever) totally effective. Better is to filter the noise as and when there is an opportunity.
All I can say is that the PhoenixNET works for me and it seems to help countless other people so it is worth having a go if you can borrow a demo one or try one with a refund option.
3 Likes
Etherregen + Eno did not work well in my system. A standard $20 D-Link switch gave me better results.
1 Like